Abstract
Background: Radiography is often used routinely by the general practitioner (GP) in knee osteoarthritis (KOA), even though the diagnosis can be made based on clinical findings. However, radiography may also be requested when serious pathology is suspected. The agreement between the radiographic and the GPs clinical diagnosis and the ability of radiography to rule out serious pathology in clinical KOA is unknown, despite that this is important to evaluate the clinical value of radiography.
Objectives: The objectives were to evaluate agreement between the radiographic and clinical diagnosis in KOA; and to describe radiographic features in patients referred from their GP with clinical KOA.
Methods: Referral forms from GPs and radiographs of 1 334 consecutive patients above 40 years not previously diagnosed with KOA were evaluated. The agreement between primary indication for radiographic referral (± clinical KOA; according to the recommendations from the European League Against Rheumatism) and the radiographic diagnosis (± radiographic KOA; Kellgren and Lawrence score ≥ 1) was estimated using Cohen's Κ statistics. Furthermore, an evaluation of radiographic features was carried out in patients with clinical KOA (n = 997).
Results: The strength of the agreement was 0.106 to 0.298 with the lowest agreement in the youngest patients and the highest in the oldest patients. Five radiographs (0.5%) revealed conditions needing further investigation or specific treatment (osteonecrosis, osteochondral lesion, fracture and subluxation).
Conclusion: In patients with clinical KOA, the radiography seems only indicated if the clinical assessment cannot rule out other diagnoses or serious pathology.
KEY MESSAGE:
There is only slight to fair agreement between the radiographic and clinical diagnosis in knee osteoarthritis (KOA) and few radiographs reveal treatment changing pathology.
Radiography in clinical KOA seems only applicable if the general practitioner's clinical assessment cannot rule out serious pathology or if non-surgical treatment is ineffective.
INTRODUCTION
Up to one third of people aged 65 or older suffers from knee pain (Citation1). Knee pain is a common reason for people to consult their general practitioner (GP) (Citation2), and in older people this pain is often due to osteoarthritis (OA) (Citation3). According to the European League against Rheumatism (EULAR) KOA can be clinically diagnosed based on the following characteristics: age > 40, usage-related knee pain, morning stiffness, functional limitation and one or more typical examination findings crepitus, restricted movement, bony enlargement (Citation4).
Hitherto, routine referral to radiography has often been used in general practice, possibly due to lack of knowledge of current diagnostic guidelines among GPs (Citation5) or for non-diagnostic reasons (Citation6–8), potentially postponing initiation of treatment due to waiting lists. However, many patients with clinical KOA do not have radiographic KOA, and the discrepancy between clinical findings and findings in radiography is well known in the scientific community (Citation9). Therefore, an evaluation of the clinical value of routine radiography in general practice is needed and should include an analysis of the agreement between the radiographic and clinical GP diagnosis and a description of the radiographic features in patients with clinical KOA.
Even though the association between clinical and radiographic KOA has been extensively evaluated (Citation9), only one study has evaluated the actual agreement between clinical and radiographic KOA finding moderate agreement (Citation10). Unfortunately, the study only evaluated a relatively small study population (n = 130, only 23 patients with clinical KOA) possibly affecting the validity of the findings (Citation11).
The other application of radiography for the GP could be to rule out serious pathology and differential diagnoses (Citation8). However, radiographic features in clinical KOA have not yet been described, even though this is important to evaluate the significance of radiography.
Therefore, the objectives were twofold: to evaluate the agreement between the radiographic (± a radiographic diagnosis of KOA) and clinical diagnosis given by the GP (± clinical KOA) in a large cohort; and to describe radiographic features in patients referred from GP with clinical KOA.
METHODS
Study design
This was a study of agreement between radiographic and clinical KOA in patients referred to radiography of their knee by their GP in the North Denmark Region in the period July 2010–December 2011. Furthermore, it evaluated radiographic features in patients with clinical KOA. The study conforms to the GRRAS guidelines for reporting reliability and agreement studies (Citation12).
Study population
Radiographs of 1 793 consecutive patients were evaluated. Inclusion criterion was GP referral for radiographic assessment. The patients were excluded if they meet one of following criteria: Age ≤ 40 (Citation4) or previously diagnosed KOA. The last criterion was evaluated by searching hospital computer records (available from 2005 and considered to be complete) using codes from the classification system outlined by the Danish Health and Medicines Authority and statements in the referral form indicating a previous diagnosis of KOA.
One thousand, three hundred and thirty-four consecutive radiographs, with indication for radiographic referral, meeting the above-mentioned criteria were included in the agreement study (only one knee per patient) while 997 patients with clinical KOA were included in the evaluation of radiographic features ().
Figure 1. Flow chart. GP, general practitioner; KOA, knee osteoarthritis; PA, psoriatic arthritis; RA, rheumatoid arthritis; GA, gouty arthritis.
The study was conducted in accordance with the Helsinki Declaration and approval was not needed according to the local ethics committee of the North Denmark Region.
Procedure
The Department of Radiology at Vendsyssel Hospital, Frederikshavn is a highly specialized unit performing approximately 45 000 assessments each year, especially serving the GPs in the region.
In Denmark, GPs use standard referral forms when referring patients for radiography. This referral form has two sections relevant for the study. The first section asks, which radiography the GP would like to be performed while the second holds a field asking for a reason for referral and a free text entry of clinical details including symptoms and primary diagnosis. Both sections are completed by the GP and are compulsory to fill out.
In the Department of Radiology, semi-flexed metatarsophalangeal posteroanterior (MTP PA) radiographs of both knees are obtained in a standing position with the feet pointing straight forward. The MTP PA view is considered to be superior to other views when assessing characteristics of radiographic KOA severity (Citation13). Furthermore, the symptomatic knee is exposed from lateral in lying position. The X-ray beam is centred at the level of the knee joint. Tube to film distance is 115 cm.
All radiographs in the Department of Radiology are assessed by two of five experienced radiologists, independently, for signs of radiographic KOA such as joint space narrowing, osteophytes, sclerosis and bone end deformities. If the two radiologists disagree, further examinations are performed. This approach ensures a high quality of the radiographic assessment.
Retrospectively, an experienced radiographic nurse (HT) seeing KOA patients daily at the hospital reviewed all GP referral forms and radiographic assessments of the study period and evaluated both clinical and radiographic KOA, using Kellgren and Lawrence grading system, based on the definitions presented below. None of the authors was involved in the radiographic referral or evaluation of the radiographs.
Evaluation of agreement
In all included radiographic assessments, the primary indication for radiographic referral was registered from the GP referral form. The indications were dichotomized into ± clinical KOA, with clinical KOA defined using EULARs recommendation number five, consisting of age > 40, usage-related knee pain, morning stiffness, functional limitation and one or more typical examination findings (crepitus, restricted movement, bony enlargement (Citation4). Whenever in doubt of whether or not the patient had clinical KOA the radiographic nurse evaluating the GP referral forms evaluated the individual referral form using EULARs list of typical symptoms of KOA, including usage-related or more persistent rest and night pain; feelings of ‘giving way;’ morning or inactivity stiffness and impaired function (Citation4).
Subsequently, all radiographs were scored as ± a radiographic diagnosis of KOA. The radiographic diagnosis of KOA was defined as the patient having a Kellgren and Lawrence score ≥ 1 on the original scale (Citation14,Citation15) since the original scale has proven to have the strongest association with knee complaints (Citation16). Furthermore, gender and age were recorded for all patients.
Evaluation of radiographic features
An evaluation of radiographic features, determined by the radiologists, were carried out in the sub-group of patients with clinical KOA as the primary indication for radiographic referral (n = 997) ().
The radiographs were divided into descriptive categories of which more than one could appear at the same time:
(1) Osteoarthritis;
(2) Joint swelling;
(3) Soft tissue calcification;
(4) Chondrocalcinosis;
(5) Joint loose bodies;
(6) Other findings; and
(7) Nothing abnormal detected.
Statistical analysis
The agreement between primary indication for radiographic referral (± clinical KOA) and the radiographic diagnosis (± radiographic KOA) was estimated using Cohen's Κ statistics (Citation17). K values can be classified as poor: < 0.00; slight: 0.00–0.20; fair: 0.21–0.40; moderate: 0.41–0.60; substantial: 0.61–0.80; or almost perfect: 0.81–1.00 (Citation17).
The agreement analysis was conducted for all 1 334 patients and the following sub-groups: men; women; young (41–60 years old); old (61 + years old). All analyses were done using SPSS Statistics (Version 19).
RESULTS
Prevalence of KOA
The prevalence of clinical and radiographic KOA in this study population is presented in . Clinical KOA was present in 997 (74%) of the patients while 686 (51%) had radiographic KOA.
Table 1. Patient characteristics.
Agreement between the radiographic and clinical diagnosis
The strength of the agreement between radiographic and clinical KOA was slight to fair (K value of 0.106 to 0.298) with the lowest agreement in the youngest patients and the highest agreement in the oldest patients ().
Table 2. Agreement (cross-tabs) between clinical KOA and radiographic KOA.
Radiographic features in clinical KOA
describes the radiographic features in patients referred from their GP for radiography due to clinical KOA (n = 997). Of the patients, 585 (58.7%) had radiographic KOA and 310 (31.1 %) of the radiographs did not show anything abnormal. Other findings consisted of non-osteophyte exostosis (n = sequelae of fractures (n = cysts (n = 5); foreign bodies (n = 4); arteriosclerosis (n = 3); patella bipartita (n = 3); osteonecrosis (n = 2); osteochondral lesion (n = 1); sequelae of osteochondral lesion (n = 1); fracture of fibula and subluxation of the proximal tibiofibular joint (n = 1).
Table 3. Radiographic features in clinical KOA (n = 997).
DISCUSSION
Main findings
This study found only slight to fair agreement between the radiographic and clinical diagnosis given by the GP and only a small proportion of the radiographs in patients with clinical KOA revealed serious, treatment changing pathology. This questions the value of routine radiography in KOA in general practice.
Strengths and limitations
Study design. The retrospective design is a limitation. A prospective design would give the possibility to control for factors attributed to the diagnosis of KOA. Both clinical and radiographic KOA were evaluated retrospectively based on GP referral forms and radiographic assessments possibly affecting data quality since the identification of patients with KOA in some cases were difficult due to missing information in GP referral forms and radiographic assessments. This could result in misclassification of some patients in the study. However, the retrospective design applied in this study gives the possibility to evaluate the actual agreement found in a clinical setting without GPs and radiologists being influenced by external factors such as awareness of study participation.
Study population. The large and presumably non age and gender skewed sample is a strength since the discordance between radiographic and clinical assessment in KOA could be due to small sample sizes and age and gender skewed cohorts (Citation9,Citation11). Hence, the agreement found in this study is based on the largest and most comparable cohort presented so far.
Radiographic and clinical KOA. Given that the diagnosis of clinical and radiographic KOA requires some interpretation, there is a potential risk of bias related to the fact that the radiographic nurse evaluated both clinical and radiographic KOA. However, since none of the authors were involved in the radiographic referral (completed by GP) or the evaluation of the radiographs (assessed by radiologist), the risk of bias related to the evaluation process was minimized. Additionally, skyline view was not included in the radiographic evaluation. This could be seen as a limitation, since the addition of skyline radiographs to posteroanterior and lateral radiographs increases the prevalence of radiographic KOA with 12% (Citation18). However, the cut-off value applied in our study for diagnosing radiographic KOA, a Kellgren and Lawrence score ≥ 1 on the original scale, is recommended to distinguish between definite/possible KOA and no KOA (Citation16). Furthermore, the definition of clinical KOA followed the recommendations from EULAR. The evidence-based definitions of radiographic and clinical KOA could, therefore, be regarded as strengths in this study.
Evaluation of radiographic features. Another strength is the evaluation of radiographic features in patients with clinical KOA. Besides a previous study describing KOA related features (Citation19), this is the first study reporting specific radiographic features, even though this is important to explore the significance of radiography as a method to rule out serious pathology and differential diagnoses in general practice (Citation8).
Interpretation
Prevalence of KOA. The study population illustrated that a higher proportion of women than men had radiographic (Citation11,Citation20,Citation21) and clinical KOA, which is in concordance with the literature. Furthermore, similar to previous studies, this study found an increase in the prevalence of KOA with increasing age (Citation11,Citation20,Citation21) and a higher prevalence of KOA in older women than any of the other sub-groups (Citation11,Citation20,Citation21).
In patients not having clinical KOA, the proportion not having radiographic KOA is around twice as big as the proportion having radiographic KOA while in patients having clinical KOA the proportion having radiographic KOA is equal to the proportion not having radiographic KOA. This suggests that GPs are better at identifying patients, who do not have radiographic OA.
Agreement between the radiographic and clinical diagnosis. The agreement between the GPs clinical diagnosis and the radiographic diagnosis was slight to fair. Only one previous study has evaluated agreement between clinical and radiographic methods of diagnosing KOA (Citation10). Toivanen et al., found a K value (95% CI) of 0.54 (0.38–0.75) comparing Kellgren and Lawrence score ≥ 2 to a clinical diagnosis of KOA. This agreement was considerably larger than the agreement found in this study. However, only 23 patients with clinical KOA were included in the agreement analysis (compared to 997 patients in our study) (Citation10). Furthermore, Toivanen et al., (Citation10) included patients previously diagnosed with KOA and applied a local set of diagnostic criteria for clinical KOA not based on international guidelines and giving the possibility to include patients with inflammatory arthritis, which could also affect the agreement (Citation22). Additionally, this study is the first evaluating the agreement using a GP given clinical diagnosis making the findings more comparable to actual general practice.
The absence of radiological established OA may lead to a different treatment in general practice (Citation6,Citation23), even though KOA management should focus on pain, sensitization and disability without worrying about type or severity of joint damage (Citation24). Moreover, a previous systematic search and summary of the literature concluded that radiographic KOA is an imprecise marker of clinical KOA and vice versa (Citation9). The proportion of patients with knee pain with radiographic KOA ranged from 15–76% while 15–81% of patients with radiographic KOA had knee pain (Citation9).
A recent study found that only 29% of GPs disagree with the statement ‘X-ray is necessary to diagnose OA’ (Citation5). However, based on our study, and since both guidelines from EULAR (Citation4) and the American College of Rheumatology (Citation25) give the possibility to diagnose KOA without radiography, the value of routine radiography in the diagnosis and management of KOA in general practice is not apparent.
Radiographic features in clinical KOA. Radiography should be clinically justified, and the result should be likely to influence patient management, also in general practice. The study illustrated that most of the radiographs did not reveal a pathology or differential diagnosis that would change what would be considered the optimal first line non-surgical treatment of patients. Only five of the 997 radiographs (0.5%) revealed conditions presumably needing a different treatment approach (osteonecrosis, osteochondral lesion, fracture and subluxation). This could indicate that radiography in general practice is applied not only due to clinical observations, but also for non-diagnostic reasons. This is supported by previous studies, indicating that GPs state reasons such as patient pressure, preservation of the doctor–patient relationship and aid to discussion on patient management for radiography referral (Citation6–8). However, the appropriateness of such applications of radiography can be challenged from existing evidence and recommendations (Citation6,Citation8). Interestingly, studies in back pain show that patients who have had radiography report more pain, lower overall health status, and consult their GP more frequently (Citation27,Citation28) giving important considerations for GPs referring patients for radiography in KOA. Furthermore, it is important to consider if the number of pathological findings can compensate for the radiation exposure from radiography and expenses related to the use of radiography. Based on our study this does not appear to be the case since the disadvantages outweigh the advantages from the routine radiography.
Implications
Radiography in patients with clinical KOA seems only applicable if the GP cannot rule out serious pathology or provide differential diagnoses based on the clinical assessment or if the initial treatment is ineffective in reducing symptoms. Multimodal, non-surgical treatment consisting of exercise, patient education, weight loss (if relevant) and pain treatment (if needed) are considered to be the first-line treatment for KOA (Citation29–31) and GPs should either initiate or refer the patient to this treatment before considering radiographic referral.
Conclusion
This study found only slight to fair agreement between the radiographic diagnosis and the clinical diagnosis given by the GP in a large study population and few radiographs in patients with clinical KOA revealed treatment changing pathology. This questions the value of routine radiography in general practice in KOA.
Notice of correction
The version of this article published online ahead of print on 30 Aug 2013 contained an error on page 1. The sentence “In patients with clinical KOA, the radiography seems to indicate only if the clinical assessment cannot rule out other diagnoses or serious pathology” should have read “In patients with clinical KOA, the radiography seems only indicated if the clinical assessment cannot rule out other diagnoses or serious pathology”. The error has been corrected for this version.
ACKNOWLEDGEMENTS
The authors would like to thank general practitioners Kirsten Dal and Søren Vinther-Nielsen for providing feedback on the manuscript.
Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for both content and writing of the paper.
REFERENCES
- Dawson J, Linsell L, Zondervan K, Rose P, Randall T, Carr A, et al. Epidemiology of hip and knee pain and its impact on overall health status in older adults. Rheumatology (Oxford). 2004;43:497–504.
- Peat G, McCarney R, Croft P. Knee pain and osteoarthritis in older adults: A review of community burden and current use of primary health care. Ann Rheum Dis. 2001;60:91–7.
- Mantyselka P, Kumpusalo E, Ahonen R, Kumpusalo A, Kauhanen J, Viinamaki H, et al. Pain as a reason to visit the doctor: A study in Finnish primary health care. Pain. 2001;89: 175–80.
- Zhang W, Doherty M, Peat G, Bierma-Zeinstra MA, Arden NK, Bresnihan B, et al. EULAR evidence-based recommendations for the diagnosis of knee osteoarthritis. Ann Rheum Dis. 2010;69:483–9.
- Smink AJ, Bierma-Zeinstra SM, Dekker J, Vlieland TP, Bijlsma JW, Swierstra BA, et al. Agreement of general practitioners with the guideline-based stepped-care strategy for patients with osteoarthritis of the hip or knee: A cross-sectional study. BMC Fam Pract. 2013;14:33.
- Bedson J, Jordan K, Croft P. How do GPs use X-rays to manage chronic knee pain in the elderly?A case study. Ann Rheum Dis. 2003;62:450–4.
- Morgan B, Mullick S, Harper WM, Finlay DB. An audit of knee radiographs performed for general practitioners. Br J Radiol. 1997;70:256–60.
- Peat G, Croft P, Hay E. Clinical assessment of the osteoarthritis patient. Best Pract Res Clin Rheumatol. 2001;15:527–44.
- Bedson J, Croft PR. The discordance between clinical and radiographic knee osteoarthritis: A systematic search and summary of the literature. BMC Musculoskelet Disord. 2008;9:116.
- Toivanen AT, Arokoski JP, Manninen PS, Heliovaara M, Haara MM, Tyrvainen E, et al. Agreement between clinical and radiological methods of diagnosing knee osteoarthritis. Scand J Rheumatol. 2007;36:58–63.
- Laxafoss E, Jacobsen S, Gosvig KK, Sonne-Holm S. Case definitions of knee osteoarthritis in 4,151 unselected subjects: Relevance for epidemiological studies: The Copenhagen osteoarthritis study. Skeletal Radiol. 2010;39:859–66.
- Kottner J, Audige L, Brorson S, Donner A, Gajewski BJ, Hrobjartsson A, et al. Guidelines for reporting reliability and agreement studies (GRRAS) were proposed. J Clin Epidemiol. 2011;64:96–106.
- Wolfe F, Lane NE, Buckland-Wright C. Radiographic methods in knee osteoarthritis: A further comparison of semiflexed (MTP), schuss-tunnel, and weight-bearing anteroposterior views for joint space narrowing and osteophytes. J Rheumatol. 2002;29: 2597–601.
- Kellgren JH, Lawrence JS. Radiological assessment of osteo- arthrosis. Ann Rheum Dis. 1957;16:494–502.
- Kellgren JH, Jeffrey MR, Ball J. The epidemiology of chronic rheumatism. Atlas of standard radiographs of arthritis. Oxford, UK: Blackwell Scientific Publications; 1963.
- Schiphof D, de Klerk BM, Kerkhof HJ, Hofman A, Koes BW, Boers M, et al. Impact of different descriptions of the Kellgren and Lawrence classification criteria on the diagnosis of knee osteoarthritis. Ann Rheum Dis. 2011;70:1422–7.
- Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics. 1977;33:159–74.
- Duncan RC, Hay EM, Saklatvala J, Croft PR. Prevalence of radiographic osteoarthritis—it all depends on your point of view. Rheumatology (Oxford). 2006;45:757–60.
- Bedson J, Jordan KP, Croft PR. A cross sectional study of requests for knee radiographs from primary care. BMC Musculoskelet Disord. 2007;8:77.
- van Saase JL, van Romunde LK, Cats A, Vandenbroucke JP, Valkenburg HA. Epidemiology of osteoarthritis: Zoetermeer survey. Comparison of radiological osteoarthritis in a Dutch population with that in 10 other populations. Ann Rheum Dis. 1989; 48:271–80.
- Felson DT, Naimark A, Anderson J, Kazis L, Castelli W, Meenan RF. The prevalence of knee osteoarthritis in the elderly. The Framingham osteoarthritis study. Arthritis Rheum. 1987; 30:914–8.
- Kinds MB, Welsing PM, Vignon EP, Bijlsma JW, Viergever MA, Marijnissen AC, et al. A systematic review of the association between radiographic and clinical osteoarthritis of hip and knee. Osteoarthritis Cartilage. 2011;19:768–78.
- de Klerk BM, Willemsen S, Schiphof D, van Meurs JB, Koes BW, Hofman A, et al. Development of radiological knee osteoarthritis in patients with knee complaints. Ann Rheum Dis. 2012;71:905–10.
- Dieppe PA. Relationship between symptoms and structural change in osteoarthritis: what are the important targets for therapy?J Rheumatol. 2005;32:1147–9.
- Hochberg MC, Altman RD, Brandt KD, Clark BM, Dieppe PA, Griffin MR, et al. Guidelines for the medical management of osteoarthritis. Part II. Osteoarthritis of the knee. American College of Rheumatology. Arthritis Rheum. 1995;38:1541–6.
- Cibere J. Do we need radiographs to diagnose osteoarthritis?Best Pract Res Clin Rheumatol. 2006;20:27–38.
- Baker R, Lecouturier J, Bond S. Explaining variation in GP referral rates for X-rays for back pain. Implement Sci. 2006;1:15.
- Kendrick D, Fielding K, Bentley E, Miller P, Kerslake R, Pringle M. The role of radiography in primary care patients with low back pain of at least 6 weeks duration: a randomised (unblinded) controlled trial. Health Technol Assess. 2001;5: 1–69.
- Jordan KM, Arden NK, Doherty M, Bannwarth B, Bijlsma JW, Dieppe P, et al. EULAR Recommendations 2003: An evidence based approach to the management of knee osteoarthritis: Report of a Task force of the standing committee for international clinical studies including therapeutic trials (ESCISIT). Ann Rheum Dis. 2003;62:1145–55.
- National Collaborating Centre for Chronic Conditions (UK). Osteoarthritis: National clinical guideline for care and management in adults.London: Royal College of Physicians; 2008.
- Zhang W, Moskowitz RW, Nuki G, Abramson S, Altman RD, Arden N, et al. OARSI recommendations for the management of hip and knee osteoarthritis. Part II: OARSI evidence-based, expert consensus guidelines. Osteoarthritis Cartilage. 2008;16:137–62.